The Information Content of Put Warrant Issues

The Information Content of Put Warrant Issues Scott Gibson a Paul Povel b Rajdeep Singh b February 2006 a Department of Economics and Finance, School of Business, College of William and Mary, Williamsburg, VA 23187. b Department of Finance, Carlson School of Management, University of Minnesota, 321 19th Avenue South, Minneapolis, MN 55455. Email: [email protected] (Gibson), [email protected] (Povel) and [email protected] (Singh). We are grateful to Sugato Bhattacharyya, Francesca Cornelli, Gustavo Grullon, Dirk Jenter, Jack Kareken, Ross Levine, Bob McDonald, Roni Michaely, Sheridan Titman, Andrew Winton, and seminar participants at the 11th annual Financial Economics and Accounting conference at Ann Arbor, MI, and at University of Minnesota and Cornell University for their helpful comments. The Information Content of Put Warrant Issues Abstract We analyze why ¯rms may want to issue put warrants, i.e., promises to repurchase their own shares at a given price in the future. We describe four alternative explanations, one of which is novel: that put warrants are issued by ¯rms that wish to signal their good future prospects to their investors (who undervalue the ¯rms in the eyes of their managers). We test the validity of the four alternative explanations, using a new, hand-collected data set on put warrant issues in the U.S. between 1993 and 1999. We ¯nd evidence that is inconsistent with three of the four explanations. Only the signaling explanation is consistent with the empirical evidence. Put warrant issuers strongly outperform their peers in the years after the put warrant issues; they enjoy valuable and improving investment opportunities, and they invest heavily. Put warrant issuers are thus very di®erent from other ¯rms with ongoing open market share repurchase programs. JEL codes: G35, G32, D82 Key Words: Put warrants; share repurchases; signaling 1 Introduction Corporate-issued put warrants give investors the right to sell shares back to the ¯rm at a future date at a ¯xed strike price. In return, the ¯rm receives an upfront tax-free cash premium. The amounts involved can be large: the most proli¯c issuer of put warrants, Microsoft Corp., collected $1.9bn between 1994 and 1999. However, put warrants expose the issuing ¯rm to potentially large liabilities, if the share price moves below the strike price. In 2000, Microsoft Corp. had potential put warrant obligations of over $11bn outstanding; during the ¯scal year 2001, the company decided to abort its put warrant program, and it spent over $1.4bn to repurchase outstanding put warrants. Thus, issuing put warrants can yield tax-free income for a ¯rm, but that brings with it potentially large and costly ¯nancial obligations. In this paper we analyze why ¯rms may want to expose themselves to this risk. We consider explanations that can be found in the media and the literature, and we develop a new explanation. We then test the validity of these explanations, using a new, hand-collected data set. Put warrants, and the reasons for issuing them, have not been systematically analyzed before. One possible explanation for their use is that put warrants allow ¯rms to synthesize an open market share repurchase: instead of going through a broker, ¯rms can use derivatives to schedule regular repurchases at certain prices (see Grullon and Ikenberry (2000)). Doing so may allow the ¯rm to reduce the transaction costs of regular share repurchases, and to avoid market imbalances if its shares are not traded in a liquid market. This explanation, which we term the transaction costs explanation, suggests that put warrant issuers are ¯rms with illiquid share markets, for example smaller ¯rms. A second possible explanation has to do with stock option compensation, and the dilution caused to existing shareholders. Some media commentators have argued that put warrants simplify share repurchases motivated by the exercise of such stock options, by us- 1 ing derivatives for the repurchases, too. In fact, some ¯rms purchased call options on their own shares. However, these were usually accompanied by the issue of an equal or larger number of put warrants, which do not link share repurchases with stock option exercise as claimed. Nevertheless, we examine the idea that put warrant issues may be more attractive for ¯rms that make heavier use of stock option compensation. The dilution explanation implies that put warrant issuers are no di®erent from other ¯rms that have ongoing share repurchase programs, except for the presence of signi¯cant option compensation programs for their executives or employees. Thus, put warrants should be issued mainly by ¯rms that o®er employees signi¯cant option compensation, such as high-tech ¯rms (see Anderson et al. (2000) and Ittner et al. (2003)). A third possibility is that ¯rms (potentially) su®ering from the \free cash flow problem" (see Jensen (1986)) issue put warrants. The agency explanation suggests that put warrant issuers are mature ¯rms, characterized by strong, stable cash flows and few or no investment opportunities. Such ¯rms may undertake bad investments, simply because managers prefer to re-invest spare cash to paying it out to shareholders. One way for managers to commit not to waste spare cash is to credibly promise to pay it out, which can be achieved with put warrants. A fourth alternative explanation is that put warrant issues signal good news to investors. The signaling explanation predicts that only ¯rms with good prospects (compared with the market's expectations) will issue put warrants, since ¯rms with poor prospects ¯nd them more costly than ¯rms with good prospects. Firms can use a variety of methods to signal good news to their investors.1 Put warrants have the bene¯t that they may be costless, or even a pro¯table way of signaling: Microsoft Corp. (and other issuers) prided themselves of having issued put warrants quarter after quarter, and not a single one was exercised. The contribution of this paper is twofold. First, we develop the signaling explanation 1See Grullon and Ikenberry (2000); Allen and Michaely (2003); Grullon and Michaely (2002); or Grullon and Michaely (2004). 2 for put warrant issues, i.e., we show theoretically how and under what conditions put warrants can be used to support a separating equilibrium. Second, we test the merits of this explanation and the alternative explanations empirically, using a new, hand-collected data set. An example can illustrate the idea of the signaling explanation. In late 1994, AT&T announced that it would team up with a local ¯rm to enter the Mexican telecommunications market in 1997, competing with Telmex (Telefonos de Mexico), the Mexican tele- phone monopoly, instead of cooperating with it. At the time, uncertainty existed about future deregulation moves of the Mexican government, and about the prospects for di®erent telecommunications markets in Mexico (long-distance service; local markets; cellular phone operations; etc.) Telmex' share price had been in decline for two weeks after the news hit. To stop this decline, its controlling shareholder, Carlos Slim Helu, announced that he would issue put warrants for Telmex stock, with a maturity of twelve months. The total potential liability if the put warrants were exercised amounted to $450m.2 The put warrant issue was meant to reassure investors that Telmex was in a strong position. Put warrants are similar to a money-back guarantee: if shareholders are not satis¯ed with the shares they purchased (because their price fell), they have the right to sell these shares back at a certain price. Clearly, such an o®er will be made only if the chances of actually having to pay up are not too large. Thus, with a su±ciently large o®er, investors should realize that the ¯rm issuing put warrants must have better prospects than expected: if the prospects were worse than investors expected, the put warrants could turn out to be very costly for the ¯rm. In other words, an appropriately structured issue can convey good news to investors, private information that would otherwise be hard for management to convey in a credible way.3 2See \Mexico's Slim Makes Big Bet On Telmex," Wall Street Journal, November 23, 1994. A similar story, featuring WMX Technologies Inc., is described in \WMX sells options in buyback strategy involving less cash," Wall Street Journal, April 11, 1994. 3In an information brochure about stock buybacks, the investment bank Salomon Brothers emphasizes the possible use of put warrants to send a positive signal to investors: \. put warrants send a positive 3 Importantly, the net costs of using put warrants to signal are zero: ¯rms that issue put warrants earn a cash premium at the time of issue, and in a separating equilibrium, the put warrants are priced fairly, so the expected cash outflow is o®set by the cash premium. In contrast, if a ¯rm with poor prospects tries to fool investors by issuing put warrants, it should expect to su®er a large and costly cash outflow when the put warrants mature, larger than the amount it collected when issuing the puts. If ¯rms with good prospects choose the right structure for their issue, then ¯rms with poor prospects will not imitate them, and investors will understand the signals and price the put warrants (and the shares) fairly. Necessary conditions for such costless (\non-dissipative") signaling have been analyzed in Brennan and Kraus (1987) and Nachman and Noe (1994). Our model focuses on put warrants, which have been issued in practice, showing that they satisfy the necessary conditions. This contrasts with other signaling models in which ¯rms use payout policy to signal inside information to outside investors. In contrast to ours, these models are based on costly signaling (see, e.g., Bhattacharya (1979), Miller and Rock (1985), Vermaelen (1984), Ofer and Thakor (1987), and Constantinides and Grundy (1989)).

Load more